There’s a big debate in the sports industry during the time of Covid-19: is it safer to play or not to play? Without much thought, it’s easy to assume “no play” is the obvious answer. No huddles, no contact, no problem. While this may be true, it isn’t realistic. Aviation would be safer, too, if planes never left the ground. Yet, business must continue. Athletes must play and airplanes must fly. A world without risk is not possible; a world with too much risk is unacceptable. To minimize risk, we must go beyond standard concepts of safety. We need to realize that actionable information is the key to minimizing risk.
The challenge with quantifying safety in our industry is that it’s nearly impossible to track a lack of outcomes. Instead, the industry is relying on abstract and incomplete information to guide decision making, and subsequently insurance premiums. So, what is the best way to determine insurance premiums? How can we determine an operator’s level of safety? Before we can answer these questions, we must understand the inadequacies of how the ‘safeness’ of an operator is widely measured.
The absence of incidents or accidents is generally used to indicate a company’s level of safety. While this can be true, it is a woefully inadequate way of gauging the safeness of an operator. As an analogy, consider a drunk driver who does not cause an accident. Does that mean he was safe? No, the danger was still there, despite the lack of a bad outcome. The lack of outcomes are not a reliable method for determining safeness.
The consistent use of hazard scoring systems is another indicator of a company’s safeness. These systems assign arbitrary values to various hazards. For example, night may be assigned a score of ‘5’ and a wet runway may be assigned a value of ‘3’. These values are then added together to get an overall score of ‘8’. Here the problem lies in assuming qualitative values can be used quantitatively. After all, we cannot claim that two 3-star restaurants are better than a 5-star restaurant simply because the sum of the two 3-star restaurants (‘6’) is greater than ‘5’. Scoring systems fail to address the risk in a meaningful way and lull users into a false sense of security.
Third-party auditors are also used to determine an operator’s level of safety. These auditing organizations provide a 2-tier or 3-tier rating system that categorizes an operator’s ‘safeness’ based on the level of conformance to an accepted standard. These third-party audits are certainly an improvement over the aforementioned methods; however, the data accumulated during the audit does not paint the complete picture. The main pitfall with audits is that they only capture the organization’s safeness during a very small snapshot of time.
Finally, participation in data sharing programs – e.g., Aviation Safety Information Analysis and Sharing (ASIAS) program, Aviation Safety Reporting System (ASRS) – is another indicator of an organization’s interest in safety. It’s important to recognize that most of these programs were not created to help individual companies; they were created to gather data for the FAA or other similar Regulatory body. While these data sharing systems have many benefits, they also have drawbacks: the data collected in these programs is inconsistent, incomplete, and difficult for operators to access.
So, how do we overcome these shortcomings to comprehensively quantify safety for the purpose of offering personalized insurance premiums? It’s easy to say we need better data analysis and information sharing. Though fundamentally important, these are not novel concepts. The multifaceted solution lies in the type of data we’re tracking and way that information is shared. Let’s dig deeper.
We’ve already established that collecting safety data is a must here, but it’s equally important to discuss the different types of data.
Take, for example, the operator who looks at the safety data and sees that they’ve had four unstable approaches in the last two weeks. If they were to act only on this data (i.e., the number of unstable approaches), they would likely conclude that pilots require remedial training. They would remind the pilots of the published stabilized approach criteria. If, on the other hand, the operator conducts a proper root cause analysis, they may have discovered that all the unstable approach events were completely unrelated – they all had different causes. Since they had different causes, they will require different remedial actions to fix. Therefore, the original proposed risk control (more training) will not fix the problem!
Consider a different example where the safety data shows a few sporadic events (e.g., unstable approach, near mid-air collision, incomplete pre-flight checks) over the past couple of weeks, but nothing seems to be trending. If the operator were to do a root cause analysis on each event, they may actually discover a common causal factor – fatigue. Only after identifying the root causes can we put in place corrective actions to avoid negative outcomes.
In order to understand the cost of safety, we need to look at more than just the cost of incidents and accidents. We need to systematically capture the impact that safety-related deviations and malfunctions have on a business. How many flight delay’s did they cause? How many flight cancellations? How much revenue was lost? How much did it hurt the organization’s reputation? By capturing these hidden costs, we can get a more complete picture regarding the true cost of safety.
If we can track the costs of a lack of outcomes – which we can, given today’s technology – then we’re really not that far away from being able to achieve objective safety analysis. As flight operations and insurance underwriters embrace information sharing, we can create industry-level safety analyses that will highlight which companies are lagging and which are leading. Not only will this process allow underwriters to accurately assess the ‘safeness’ of an operator compared to their peers, it will allow flight operations to justify future safety investments, thrusting the benchmark for safety into unchartered territory.
So, are personalized insurance premiums possible? The more we use data to provide objective context to the safety standing of our organization and can benchmark ourselves against comparable industry metrics, the opportunity for this type of change is legitimate.
Co-Authors: Steve Bruneau and Madeline Young
As first published in the AIA Binder Magazine Fall 2020 edition.
Published: 19th October 2021
Lead Photo credit: Photo by Nick Romanov on Unsplash
;
